Performance Analysis of Channel Model Mismatch for RIS-aided Positioning Systems

It is clear that the next-generation of communication systems will move towards wireless positioning systems based on high frequency and large bandwidth. Moreover, since reconfigurable intelligent surfaces (RIS) have good commercial prospects, RIS-aided indoor near-field (NF) positioning will be a hot topic for future research. Therefore, it is no longer reasonable to use the common stationary narrowband far-field (FF) channel model and FF path loss (FPL) model in the NF region of RIS. And, it is necessary to consider the effect of channel spatial non-stationarity (SNS), beam squint effect (BSE), NF path loss (NPL) model and NF spherical wave model (SWM) to build a true channel model. If the FF channel model instead of the true channel model is utilized in the positioning process, the problem of channel model mismatch occurs, and the positioning performance will be degraded. In this work, we evaluate the positioning error bound (PEB) of a true channel model by the Cramer-Rao bound (CRB), the PEB of a mismatched channel model by the misspecified CRB (MCRB), then investigate the performance loss by comparing the CRB and MCRB. Simulations show that the RIS-aided positioning performance is degraded by using the mismatched channel model. Also, in the NF region NPL and SWM have the large affect on model mismatch, and SNS has the smallest affect. In FF region, the impact of BSE is more significant.